[Webkit-unassigned] [Bug 39557] Full screen doesn't work for video elements

Mon May 24 08:50:52 PDT 2010

https://bugs.webkit.org/show_bug.cgi?id=39557

--- Comment #21 from Jer Noble <jer.noble at apple.com>  2010-05-24 08:50:52 PST ---
(In reply to comment #18)
> Are we sure no platform will throw a divide by 0 exception? If so, then it's fine. I'm just not aware of code elsewhere in WebKit that allows these sorts of shenanigans.

Float div by 0 will always return either +INF or -INF.

> We never use this API in GraphicsLayer. We have to do lots of logic to flip Y and do other conversions so we always do that at the higher levels anyway. I had a couple of places where it was done in WKCACFLayer. Because my position is always(0,0), I could set the anchor point to (0,0) on creation and then just return setFrame with setBounds. If your anchor point never changes (and I assume that's the case) you can simply set the anchor point to (0,0) and then call setBounds(CGRectMake(0,0, rootBounds.size.width, rootBounds.size.height); setPosition(rootBounds.origin).

Right, I could, assuming I want an anchor point of (0,0).  But how is all of that less error prone than just calling setFrame()?

> I removed setFrame to avoid needing to deal with its complexities in my tiled layer logic. It's long been the source of confusion and bugs, and since it is used so rarely, I felt (and still do) that avoiding its use is the best approach.

If anything, setFrame() is less confusing than setPosition(), setAnchorPoint(), and setBounds().  It has valid use cases, and like any API, its invalid ones.  But if it's exposed by CALayer/CACFLayer, I don't see why it should be disallowed in WKCACFLayer.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.